INSPECTION OF RECOVERY BOILERS W. B. A. (Sandy) Sharp SharpConsultant, Columbia, Maryland, U.S.A. TAPPI Kraft Recovery Short Course St. Petersburg, Florida, January 7-10, 2008 OUTLINE Purpose of inspections Inspection methods Scope and frequency of inspections 1
PURPOSE OF INSPECTIONS Verify that critical control and safety systems are operating properly ESP test Drum level and low solids trip tests Safety valve tests Instrument function tests Instrument calibration tests Verify integrity of pressure parts and evaluate needs for maintenance and/or repair Routine TYPES OF INSPECTION e.g. walkdowns Scheduled Major inspections at planned shutdowns Opportunistic When boiler becomes available for inspection for other reasons 2
AUDIT PROGRAMS USEFUL TO EVALUATE PERFORMANCE Operational performance Personnel organizations, progression, training Operator procedure checklists Boiler and auxiliary system inspections Corrosion inspections Rectification of non-compliant items Audit scope needs regular review 3
WHY DO WE NEED THESE EXPENSIVE CORROSION INSPECTIONS? Must prevent critical leaks Scheduled repairs much less expensive than unscheduled Cannot predict corrosion rates from boiler operation parameters CAUSES OF CRITICAL LEAKS Attachment Welds 41% Other 12% Butt Welds 10% Ops & Maint. Problems 15% Corrosion Thinning 22% 4
INSPECTION METHODS (1) Inspect visually To identify potential problems Measure tube thickness To detect thinning problems Find fireside cracks That could propagate into leaks Find waterside fissures That could propagate into leaks Evaluate waterside deposits That could cause overheating 5
INSPECTION METHODS (2) Visual inspection (every scheduled shutdown) by plant workers by outside experts to focus NDT particularly important with composite tubes every port, every spout, every outage Don t rely on journeyman NDT technicians to make visual inspections INSPECTION METHODS (3) Measuring tube wall thickness (every scheduled shutdown) Carbon steel tubes (total wall) digital UT oscilloscope UT internal eddy current or UT in gen. bank tubes automated UT for near-drum scanning Composite tubes (protective stainless layer) magnetic lift-off gauge eddy current instrument measure only at air ports, smelt spouts, smelt line, or if DWD very high 6
INSPECTION METHODS (4) Look for fireside cracks in areas subject to thermal shock, esp. on composite tubes, with PT Surface preparation is critical (TIP 0402-30) Measure crack depth by exploratory grinding (or shear wave UT) Greater risks in air port tubes than in floor tubes Look for waterside SAC in susceptible areas with RT Beam must aim down, not across, fissures Analyze waterside deposits INTERPRETING INSPECTION DATA You will not find critical defects unless you look where they are and use a method that can find them Judge fitness for service (consequences of discovered defects) using jurisdictional and insurance rules National Board Inspection Code (ASME in some states) skills in fracture mechanics API/ASME FFS-1 7
8
MINIMIZING ERRORS IN TUBE THICKNESS MEASUREMENTS accuracy and precision Accurate but not precise Precise but not accurate ERRORS IN TUBE THICKNESS MEASUREMENTS Standard error in precision 0.005-0.007 Standard error in accuracy 0.004-0.006 Combined standard error about 0.007 so check unexpected readings 9
SINGLE POINT RATE CALCULATIONS ARE NOT ACCURATE 0.225 Measured wall thickness 0.200 Year 1 measured thickness 0.200 +/- 0.007 Year 2 measured thickness 0.197 +/- 0.007 +0.011 /yr? True Rate -0.003 /yr -0.017 /yr? 0.175 TO MINIMIZE ERRORS IN TUBE THICKNESS MEASUREMENTS Use only technicians pre-qualified by TIP 0402-21 Recalibrate UT instruments every 100 readings, or every 15 minutes - repeat measurements if calibration drifted Check unexpected measurements - use different technician, different instrument 10
11
SCOPE AND FREQUENCY OF RECOVERY BOILER INSPECTIONS Alert observation during normal operations Systematic inspection during outages: Visual inspection and appropriate NDT to find critical defects Thickness measurements to estimate remaining life No national standards for scope, methods, frequency Develop an individual inspection plan for each boiler (5-year) SCOPE AND FREQUENCY OF RECOVERY BOILER INSPECTIONS Useful documents: AF & PA Recovery Boiler Manual, 1992 NACE (composite tubes), 1992 TAPPI (UT surveys), TIP 0402-18, 2001 (Technician qualification), TIP 0402-21, 2002 ASTM E543 (qualification of NDT contractors) ASNT TC-1A, current edition API/ASME FFS-1 Fitness-for-Service code (2007) 12
ESTABLISHING INSPECTION SCOPE AND FREQUENCY Three approaches have been taken: 1. Thickness measurements at predetermined grid locations 2. Thickness measurements based on risk of critical leaks 3. Visual observations, damage-appropriate NDT and corrosion rate calculations 1. PREDETERMINED GRID LOCATIONS No established standards for grid size or frequency wide diversity in practice Increasing number of thickness readings generally does not increase probability of detecting critical flaws Tailor inspection plans according to historical behavior this unit Problems found in similar boilers 13
1. PREDETERMINED GRID LOCATIONS Use grid readings to: estimate average corrosion rates estimate remaining life flag thinned areas Supplement grid readings with smaller grids or scanning: in areas known to be critical in areas found to be thin Calculate number of readings required to establish thinning rate in different areas of the boiler 1. PREDETERMINED GRID LOCATIONS Remember grid methods will not detect: fireside cracking of composite tubes air port balding of composite tubes thinning caused by sootblowers near-drum thinning of generating bank tubes fireside cracking at attachment welds and at headers 14
1. PREDETERMINED GRID LOCATIONS Grid methods will also not detect: thinning next to tube-to-tube membrane localized thinning at ports and at the smelt line cold-side corrosion of non-membrane water wall tubes waterside corrosion fatigue cracking (SAC) excessive waterside deposits 15
2. RISK OF CRITICAL LEAKS 32% in lower water wall, + 8% in floor + 6% at spouts 15% in upper water wall 12% in furnace screen 18% in generating bank 2. RISK OF CRITICAL LEAKS Relative grid sizes: Floor 4 x 4 Lower water wall 1 x 1 Upper water wall 2 x 2 Generating bank 2 x 2 Roof 2 ¼ x 2 ¼ Furnace screen 2 2/3 x 2 2/3 Air port tubes scan 100% Why is this a bad approach? 16
2. RISK OF CRITICAL LEAKS Critical leaks show what careless inspectors have missed, not what careful inspectors have found So, since most leaks occur at defects created when the boiler was built, use critical leak data for planning QA programs rather than for planning inspections 17
3. CORROSION RATES Establish scope and frequency of grid measurements to obtain statistically significant estimates of corrosion rate in each part of the boiler To estimate rates with fewest measurements, measure thickness at same exact locations Typically at least 100 measurements must be repeated in each part of the boiler on at least 4 occasions 3. CORROSION RATES Computer generated thickness maps show patterns of thinning useful to refine grid sizes Computer programs can: calculate thinning rates estimate remaining life indicate the probable accuracy of these predictions Don t try to calculate corrosion rates at individual points 18
19
SCOPE AND FREQUENCY RECOMMENDATIONS Use expert visual inspectors to seek critical flaws Use grid measurements to estimate corrosion rates Do additional NDT in areas: of concern from previous inspections of concern to current inspectors of concern because of findings elsewhere in the industry KEYS TO SAFE, LOW MAINTENANCE OPERATION Quality assurance to avoid building in defects during design and construction Careful operation by well-trained operators Regular, systematic, and well-documented inspection Planners and Inspectors aware of new inspection technology and of problems found elsewhere (BLRBAC, TAPPI, NACE) Timely maintenance with accountability 20
CONCLUSIONS AF & PA manual contains useful inspection planning checklists Tailored NDT plans can find thinning, cracking and internal defects Use expert visual inspectors to direct the search for thinnest tubes and critical flaws Because of (0.007 ) errors in individual thickness readings, use UT data (only) to establish corrosion patterns, project rates and schedule repairs/ rebuilds 21